Taylor expansions of solutions of stochastic partial differential equations

The solutions of parabolic and hyperbolic stochastic partial differential equations (SPDEs) driven by an infinite dimensional Brownian motion, which is a martingale, are in general not semi-martingales any more and therefore do not satisfy an Itô formula like the solutions of finite dimensional stochastic differential equations (SODEs). In particular, it is not possible to derive stochastic Taylor expansions as for the solutions of SODEs using an iterated application of the Itô formula. However, in this article we introduce Taylor expansions of solutions of SPDEs via an alternative approach, which avoids the need of an Itô formula. The main idea behind these Taylor expansions is to use first classical Taylor expansions for the nonlinear coefficients of the SPDE and then to insert recursively the mild presentation of the solution of the SPDE. The iteration of this idea allows us to derive stochastic Taylor expansions of arbitrarily high order. Combinatorial concepts of trees and woods provide a compact formulation of the Taylor expansions. Taylor expansions, stochastic partial differential equations, SPDEs, strong convergence, stochastic trees ∗Institute of Mathematics, Johann Wolfgang Goethe-University, D-60054 Frankfurt am Main, Germany, E-mail: [email protected]